In [10]:
trainer._load_dataset()
trainer.draw_dataset()
Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21)
%load_ext autoreload
%autoreload 2
import os
import warnings
warnings.filterwarnings('ignore')
from CellTracker.unet3d import TrainingUNet3D, unet3_a
%matplotlib inline
Using TensorFlow backend.
trainer = TrainingUNet3D(noise_level=100, folder_path=os.path.abspath("./unet_01"), model=unet3_a())
Made folders under: /home/wen/Projects/3DeeCellTracker-master/Code/v04 Following folders were made: unet_01/train_image unet_01/train_label unet_01/valid_image unet_01/valid_label unet_01/models
trainer._load_dataset()
trainer.draw_dataset()
Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21) Load images with shape: (512, 1024, 21)
trainer._preprocess()
trainer.draw_norm_dataset()
Images were normalized Images were divided Data for training 3D U-Net were prepared
trainer.draw_divided_train_data()
trainer._train()
Epoch 1/1 60/60 [==============================] - 64s 1s/step - loss: 0.6214 - val_loss: 0.4579 val_loss at step 1: [0.45789238313833874]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.4976 - val_loss: 0.2868 val_loss updated from [0.45789238313833874] to [0.28677449954880607]
Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.3795 - val_loss: 0.1639 val_loss updated from [0.28677449954880607] to [0.1638797413971689]
Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.2848 - val_loss: 0.1640 Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.2236 - val_loss: 0.1154 val_loss updated from [0.1638797413971689] to [0.11539153258005778]
Epoch 1/1 60/60 [==============================] - 69s 1s/step - loss: 0.1660 - val_loss: 0.0864 val_loss updated from [0.11539153258005778] to [0.08638463914394379]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.1409 - val_loss: 0.1554 Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.1063 - val_loss: 0.0966 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0940 - val_loss: 0.0419 val_loss updated from [0.08638463914394379] to [0.04194509838190344]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0740 - val_loss: 0.0481 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0680 - val_loss: 0.0476 Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.0569 - val_loss: 0.0344 val_loss updated from [0.04194509838190344] to [0.03438990149233076]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0492 - val_loss: 0.0257 val_loss updated from [0.03438990149233076] to [0.02566126449447539]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0490 - val_loss: 0.0212 val_loss updated from [0.02566126449447539] to [0.021164056172387466]
Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.0410 - val_loss: 0.0171 val_loss updated from [0.021164056172387466] to [0.01711921236063871]
Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.0391 - val_loss: 0.0198 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0390 - val_loss: 0.0269 Epoch 1/1 60/60 [==============================] - 68s 1s/step - loss: 0.0377 - val_loss: 0.0164 val_loss updated from [0.01711921236063871] to [0.0164290853879518]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0345 - val_loss: 0.0170 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0309 - val_loss: 0.0149 val_loss updated from [0.0164290853879518] to [0.01489207438296742]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0320 - val_loss: 0.0246 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0291 - val_loss: 0.0215 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0308 - val_loss: 0.0132 val_loss updated from [0.01489207438296742] to [0.01320272834143705]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0271 - val_loss: 0.0144 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0255 - val_loss: 0.0123 val_loss updated from [0.01320272834143705] to [0.012342301223220097]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0260 - val_loss: 0.0144 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0214 - val_loss: 0.0146 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0257 - val_loss: 0.0130 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0249 - val_loss: 0.0123 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0243 - val_loss: 0.0105 val_loss updated from [0.012342301223220097] to [0.010523302386799414]
Epoch 1/1 60/60 [==============================] - 66s 1s/step - loss: 0.0235 - val_loss: 0.0112 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0208 - val_loss: 0.0127 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0227 - val_loss: 0.0116 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0218 - val_loss: 0.0123 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0210 - val_loss: 0.0098 val_loss updated from [0.010523302386799414] to [0.009816732361084886]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0215 - val_loss: 0.0104 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0184 - val_loss: 0.0120 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0219 - val_loss: 0.0140 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0185 - val_loss: 0.0179 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0210 - val_loss: 0.0109 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0187 - val_loss: 0.0083 val_loss updated from [0.009816732361084886] to [0.008299452740983624]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0178 - val_loss: 0.0106 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0194 - val_loss: 0.0131 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0155 - val_loss: 0.0102 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0182 - val_loss: 0.0086 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0177 - val_loss: 0.0097 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0177 - val_loss: 0.0093 Epoch 1/1 60/60 [==============================] - 66s 1s/step - loss: 0.0192 - val_loss: 0.0091 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0156 - val_loss: 0.0093 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0180 - val_loss: 0.0123 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0199 - val_loss: 0.0105 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0167 - val_loss: 0.0092 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0182 - val_loss: 0.0113 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0156 - val_loss: 0.0091 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0170 - val_loss: 0.0096 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0137 - val_loss: 0.0089 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0144 - val_loss: 0.0120 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0160 - val_loss: 0.0102 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0138 - val_loss: 0.0090 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0169 - val_loss: 0.0081 val_loss updated from [0.008299452740983624] to [0.008112046408415254]
Epoch 1/1 60/60 [==============================] - 66s 1s/step - loss: 0.0132 - val_loss: 0.0080 val_loss updated from [0.008112046408415254] to [0.007981161577239012]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0148 - val_loss: 0.0075 val_loss updated from [0.007981161577239012] to [0.007481191782668854]
Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0178 - val_loss: 0.0094 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0164 - val_loss: 0.0082 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0151 - val_loss: 0.0084 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0129 - val_loss: 0.0096 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0154 - val_loss: 0.0099 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0130 - val_loss: 0.0101 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0130 - val_loss: 0.0092 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0128 - val_loss: 0.0086 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0120 - val_loss: 0.0082 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0143 - val_loss: 0.0084 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0108 - val_loss: 0.0133 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0160 - val_loss: 0.0105 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0141 - val_loss: 0.0077 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0147 - val_loss: 0.0089 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0140 - val_loss: 0.0080 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0118 - val_loss: 0.0095 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0128 - val_loss: 0.0087 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0122 - val_loss: 0.0093 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0128 - val_loss: 0.0098 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0128 - val_loss: 0.0086 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0110 - val_loss: 0.0091 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0130 - val_loss: 0.0094 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0106 - val_loss: 0.0156 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0126 - val_loss: 0.0088 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0128 - val_loss: 0.0086 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0104 - val_loss: 0.0080 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0200 - val_loss: 0.0109 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0121 - val_loss: 0.0087 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0121 - val_loss: 0.0084 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0104 - val_loss: 0.0092 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0110 - val_loss: 0.0102 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0112 - val_loss: 0.0112 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0105 - val_loss: 0.0093 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0122 - val_loss: 0.0120 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0099 - val_loss: 0.0097 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0109 - val_loss: 0.0079 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0112 - val_loss: 0.0075 Epoch 1/1 60/60 [==============================] - 67s 1s/step - loss: 0.0098 - val_loss: 0.0089
trainer._select_weights(step=61)